A

An enhaneed-stabiI Ity thin solenoid magnet design Is presented. The details of the high purity aluminum stab11 zed conductor are discussed. The design details of a special cryostat constructed for conductor evalua-tion Is presented, and the aluminum alloy NbTi super- conductor under procurement Is described.


Choosing
(3) Currently, several thin radiation-transparent solenoid magnets 1 » 2 » 3 are under construction for use in particle detection systems at col Iiding-beam particle accelerators. Magnet thickness Is denoted by radiation length \, .so that If a. total.thickness of 0.5 A were m desired, then 6.7 cm of copper, or 0.9 cm of stainless steel, or 4.5 cm of aluminum could be used In the construction of the magnet col Is and cryostat.
The required economy of materials necessitates the abandonment of conventional cryostable conductors with large Cu:NbTI ratios, and novel approaches to magnet stability and safety have been devised. Cryosfat design and coil cooling have also been optimized for as thin a total package as possible.
At LBL 1 , to protect the high current density conductor In the event of a quench, a low resistivity aluminum bobbin is Inductively coupled to the colt so that at the onset.of quench the current rapidly leaves the conductor and dissipates the magnet stored energy safely in the bobbin.
At SacIay 2 , high purity.aluminum directly cosoldered onto the superconducting composite provides a shunt to protect the superconductor In the event of a quench.
Codling is achieved with a tube colled (or applied serpentine fashion) onto the coll, filled with circulating helium at * 4.2 K.

CONDUCTOR DESIGN
The LBL approach precludes rapid charging of the magnet, due to the resulting massive heating of the high conductivity bobbin. Choosing a conductor directly stabilized with high-purity aluminum, we find the minimum amount of aluminum required for safety in the event of a quench from the general results of P. . Eberhard In this case, the magnet has diameter 3 m, length 4 m, and generates 1.5 T central field. Since A (NbTI) = 1.7 x I0" 6 m 2 , we have A (aluminum) =3.4 x I0" 5 m 2 . Since for 1.5 T central field, 12 kA/cm current per unit length of coll Is required, a 5000 A conductor is 0.40 cm wide, allowing for 0.02 cm turn^-to-turn Insulation. Thus, the normal metal Is 0.85 ca thick radially. Detailed calculations of a col I with these parameters, using the program QUENCH 5 and the measured quench velocity from the Saclay test coll 2 , show the ratio r to b-3 very conservative with respect to 6 max . Expression CD is retained so that scaling to other magnets is straightforward.

COIL SUPPORT
To safely maintain the aluminum resistivity as desired, the hoop stress should not cause the conductor strain to exceed 0.02? 6 .
The magnetic pressure at 1.5 T Is 130 psi so the cot I Is supported with an aluminum alloy structure of thickness Since a hollow conductor is proposed, an equivalent amount of metal Is formed into a Jacket which dads the finished conductor. In Fig. I is a sketch of the proposed conductor. The superconducting wires are cabled around a tube of SI00 or EEE aluminum alloy CRRR = 60). A thick ribbon of high-purity aluminum is rolled around this and compacted on with a strip of high strength alloy rolled Into place. The outer Jacket Is thick enough to bear the hoop load of the magnet.

COIL WINDING
The conductor is wound on a fiberglass-epoxy bobbin 0.5 cm thick, and a 0.5 cm thick layer of epoxy fiberglass Is applied to the outside of the coll. The outer layer of fiberglass !s applied with a wet I ay up process, and care Is taken to squeeze out all excess epoxy as the conductors are locked Into position. The filament winding technique can control thermal motion so that axial and radial shrinkage match the conductor upon cool down.
This type of coil fabrication, without preload In the coll winding during construction. Is seen to minimize the coll thickness since a thick metal bobbin Is not required to support the winding preload. Such a winding concept will greatiy benefit In stability If the coolant, Instead of circulating in a lube outside the coll, circulates Inside the conductor Itself.

An aluminum honeycomb-type vacuum vessel and a thin aluminum liquid nitrogen-cooled shield contribute 0.16 A. This number Increases If a more conservative vacuum vessel made of extruded aluminum panel is used. 7
With caref u i design and construct I on. the cryostat can reduce the total heat loac! to the coil to = 10 W. Use Is made of epoxy-fibergIass support struts, multilayer Insulation* and an 80 K shield on both sides of ttecoll. Figure 2 Is a plot of overal I magnet thickness, as. a function of field and radius, for a coil and cryostat incorporating these concepts.

5N Aluminum
Al/NbT. so that a great deal of condutor enthalpy is available Immediately to absorb a local disturbance. On this time scale the current never leaves the superconductor due to the extremely small magnetic diffuslvlty time constant of high-purity aluminum, e.g., a 0.5 sec.

EXPERIMENTAL DEVELOPMENT
• A special cryostat. has been constructed to measure the stability properties of the proposed conductor (Fig. 3). Subcooled liquid helium (up to 4.0 gm/sec cm 2 ) can be continuously supplied to a test section of conductor or a model coil. Thermocouples and pressure taps are available to monitor local conditions In the conductor, and a capacitor technique Is being considered to monitor fluid quality at the Inlet to the test conductors. Since flow conditions In the conductor are to be nearly stagnant, it is desirable to study the transient response of a length of the conductor to a local heat pulse.

CONDUCTOR STABILITY
This conductor can be made to appear locally cryostable. I.e., sufficient internal perimeter can be obtained so that for a choice of he-it transfer coefficient (depending on the choice of helium coolant, e.g., supercritical,.subcrltical, or even superfluid) the ohmlc heating In the composite Is completely transferred to the coolant. However, It Is easy to show that If sufficient coolant flow Is provided for classic

Fig. 3. Cryostat with Heat Exchanger and 5000 A Leads
Delivery of about 5000 m of 1100 aluminum alloy stabilized NbTI superconductor is expected in the near future. Short samples of such material exhibit good stability in contrast with 5056 alloy aluminum stabilized NbTI previously tested 6 at ANL. The overall monolithic conductor is to have 1.4 mm diameter with NbTI filament diameter 25 wm. The critical current of the conductor Is to be 1500 A at-#3T f 4.2 K, 0.1* strain.